A fundamental question faced by physicians treating acute decompensated heart failure (ADHF) is 'When has treatment worked sufficiently for safe discharge, and who requires further treatment?' Patients with ADHF have a high incidence of morbidity and mortality. Current guidelines for emergency department (ED) and hospital disposition of patients with ADHF are based on limited empirical evidence. This creates clinical uncertainty regarding disposition, leads to prolonged hospitalizations, higher costs and increased resource consumption. Conducting a project to answer this question, in combination with carefully planned didactics and mentoring from distinguished researchers, will allow me to develop as an independent clinical investigator with a primary focus on decision making in patients with ADHF. The ED is the portal of entry for the majority of ADHF admissions and presents fertile ground for prospectively studying decision making. My specific aims are: 1) to further develop knowledge in clinical research, advanced modeling, and decision sciences to advance my development as an independent clinical investigator and 2) to develop a prediction rule from readily available clinical data that helps physicians identify ADHF patients for whom early discharge from the ED and hospital is safe after treatment is initiated. In recognition of the importance of the ED in ADHF care, the NHLBI is supporting our observational cohort study of 1800 ED patients to develop a prediction rule for initial risk-stratification (STRATIFY- 1R01HL088459). To complete my aim, I will collect additional data on a subset of 500 patients from STRATIFY whose baseline data is collected within 1 hour of ED therapy. In addition to the baseline data from STRATIFY, the current proposal will collect clinical and laboratory data 2-4 hours, 12-24 hours and 96 hours after treatment is begun. We will also interview patients to measure social and behavioral variables. We will use this additional data to develop a second prediction rule to assist physicians in identifying patients safe for early ED and hospital discharge after therapy is begun. This project is an ideal vehicle for practical application of didactic knowledge gained, offers an intensive clinical research experience, and promises to generate extensive preliminary data to fully expand a long-term research program. My clinical research, didactic program and advanced biostatistical and modeling coursework will be supervised by a highly qualified, multi-disciplinary team of talented individuals, Building from my completed Outcomes Research fellowship and previous research experience, this proposal will allow me to mature as a clinical investigator, develop a prediction rule for decision making in ADHF, and engender evidence- based research in a disease process with tremendous healthcare system implications. (End of Abstract)